Pierre Divenyi

Decline of speech understanding and auditory thresholds in the elderly

A group of 29 elderly subjects between 60.0 and 83.7 years of age at the beginning of the study, and whose hearing loss was not greater than moderate, was tested twice, an average of 5.27 years apart. The tests measured pure-tone thresholds, word recognition in quiet, and understanding of speech with various types of distortion (low-pass filtering, time compression) or interference (single speaker, babble noise, reverberation). Performance declined consistently and significantly between the two testing phases. In addition, the variability of speech understanding measures increased significantly between testing phases, though the variability of audiometric measurements did not. A right-ear superiority was observed but this lateral asymmetry did not increase between testing phases. Comparison of the elderly subjects with a group of young subjects with normal hearing shows that the decline of speech understanding measures accelerated significantly relative to the decline in audiometric measures in the seventh to ninth decades of life. On the assumption that speech understanding depends linearly on age and audiometric variables, there is evidence that this linear relationship changes with age, suggesting that not only the accuracy but also the nature of speech understanding evolves with age.

Discrimination of the Silent Gap in Two‐Tone Sequences of Different Frequencies

Pairs of two‐tone sequences were presented in such a way that listeners could adjust the time interval separating the onsets of the two tones in the second pair, so that the gap or rhythmic repetitiveness appeared to be the same in the second as in the first pair. The pairs consisted of two 20‐msec tone bursts with 2.5‐msec rise time and 10‐msec fall time. The time interval separating the onsets of the tones in the first, or standard, pair was 40 msec. Responses were observed as a function of the frequency of the fourth tone, the standard pair being of one frequency (1 kHz), while the variable pair began at the same frequency but ended with a different one. As the frequency difference increases up to 250 Hz, the time interval in the variable pair for apparent equality of gap decreases, as does the variability in judgments. Beyond a difference of 250 Hz, both quantities increase. These results are consistent with those of earlier time discrimination experiments [P. L. Divenyi, J. Acoust. Soc. Amer. 49, 127 (1971)] and may relate to the time required to shift the listening band from one frequency to another.

Decreased ability in the segregation of dynamically changing vowel-analog streams: a factor in the age-related cocktail-party deficit?

Pairs of harmonic complexes with different fundamental frequencies f0 (105 and 189 Hz or 105 and 136 Hz) but identical bandwidth (0.25–3 kHz) were band-pass filtered using a filter having an identical center frequency of 1 kHz. The filters center frequency was modulated using a triangular wave having a 5-Hz modulation frequency fmod to obtain a pair of vowel-analog waveforms with dynamically varying single-formant transitions. The target signal S contained a single modulation cycle starting either at a phase of −π/2 (up-down) or π/2 (down-up), whereas the longer distracter N contained several cycles of the modulating triangular wave starting at a random phase. The level at which the target formants modulating phase could be correctly identified was adaptively determined for several distracter levels and several extents of frequency swing (10–55%) in a group of experienced normal-hearing young and a group of experienced elderly individuals with hearing loss not exceeding one considered moderate. The most important result was that, for the two f0 differences, all distracter levels, and all frequency swing extents tested, elderly listeners needed about 20 dB larger S/N ratios than the young. Results also indicate that identification thresholds of both the elderly and the young listeners are between 4 and 12 dB higher than similarly determined detection thresholds and that, contrary to detection, identification is not a linear function of distracter level. Since formant transitions represent potent cues for speech intelligibility, the large S/N ratios required by the elderly for correct discrimination of single-formant transition dynamics may at least partially explain the well-documented intelligibility loss of speech in babble noise by the elderly.

Identification of the Temporal Order of Three Tones

Monaural listeners were asked to identify permutations of a given set of three tones which differed only with respect to their frquency. In one set of experiments the duration of each tone was fixed at 10 msec, the frequency of the middle tone remained constant at 1 kHz, and the frequency difference, Δf, between the lowest and the highest tones was varied from 100 to 1500 Hz, symmetrically (log) around 1 kHz. Overall identification performance, poor at the smallest Δf value, improved when Δf was increased to 227 Hz but was unaffected by further increase in Δf. In another set of experiments the duration per tone was variable (2 to 20 msec) while the tonal frequencies were held constant (Δf=227 Hz, symmetrical around 1 kHz). From the resulting psychometric functions, one could determine tone durations necessary for identification thresholds. Across all patterns this threshold was found to be between 7 and 13 msec, depending on the observer. A feature analysis of the data shows that the terminal frequency is...

Speech acoustics, phonetics, and phonology unified by an optimal dynamic model of sound communication

A book by us (Speech: A dynamic Process, de Gruyter, 2017) presents an optimal communication model for the production of sounds through dynamic deformation of a simple tube using the criterion that a maximum contrast between consecutive sounds must be achieved by expending a minimum effort. Its output corresponds well to fundamental articulatory speech data (places of articulation, F1-F2-F3 trajectories). When driven by a periodic source, the model generates natural-sounding vowels and, through constricting the tube at distinctive regions, also stop consonants. From this model’s viewpoint, phonetic systems of vowels and consonants arise as the direct consequence of production dynamics capable of producing essentially any vowel or stop consonant. The good correspondence between characteristics of the model and those of speech leads us to conclude that (1) the model is explanatory, (2) because it is acoustically optimal so must be speech production, and (3) an acoustically anchored phonology will derive fro...

Identification of the temporal order in three‐tone patterns embedded in eight‐tone sequences

Highly trained listeners identified three‐tone patterns that were part of an eight‐tone sequence. The patterns constituted the six random permutations of three temporally contiguous tones composing a 1/3‐octave span (890, 1000, 1118 Hz). The five nonpattern “background” tones were randomly chosen, in each trial, from a 2/3‐octave range. Three parameters were investigated: duration of each component tone, temporal position of the pattern within the sequence, and the band from which the background tones were chosen. Subjects had no difficulty identifying the patterns even at very short durations (18 msec/tone), whenever the patterns occupied initial or final position in the sequence, or whenever the frequency range of the background was remote from that of the pattern. However, in those conditions where the pattern was in the middle of the sequence, the average duration for threshold performance (d′ = 1.0) was longer: about 90 msec when the background‐frequency range enclosed that of the pattern, about 65 m...

Identification of the Temporal Order of Three Tones: Effect of an Added Tone

Experienced and musically trained monaural listeners had to identify the order of three 20‐msec tones whose frequency covered a 1/3‐octave range (891, 1000, 1118 Hz) and which were followed by a fourth tone of variable frequency. The listeners were instructed to ignore this added tone. Overall identification performance was relatively unaffected when the frequency of this tone was far from the range of pattern frequencies, and was most affected when the frequency of the added tone was slightly (up to 1/3 octave) higher than the highest frequency in the pattern. When the added tone and one of the three pattern tones had the same frequency, the patterns in which the first tone coincided with the added tone became less identifiable. Results of another experiment showed that lengthening of the added tone actually improved identification of the three‐tone permutations.

Discrimination of the Silent Gap in Two‐Tone Sequences of Different Frequencies. II Dichotic Case

Two tone bursts, one to each ear, were separated by an interval, t1. Observers had to adjust the time interval, t2, between two subsequent tone bursts so as to match that interval. Each 86‐dB‐SPL burst was 20 msec long and was shaped with a 2.5‐msec rise time and a 10‐msec fall time. The leading edges of the standard tone burst pair were 40 msec apart. The frequency of the first three bursts, f1, was the same whereas that of the forth burst was f1−Δf. The geometric mean of the two frequencies was held constant at 1 kHz. As in the monaural case [Divenyi and Hirsh, J. Acoust. Soc. Amer. 52, 166(A) (1972)], the adjusted duration of t2 becomes 3–8 msec shorter than t1 as Δf increases to the third octave. This discrepancy is all but eliminated when the intensity of the fourth tone burst is lowered by 30–40 dB. [Supported by NINDS Grant No. NS03856.]